Diffusion, development, and democracy, 1800-1999

While a trend of growth in democratization over the past two centuries has been generally observed, it is the remarkable growth in the democratization of the world over the past 30 years that has truly captured the imagination of social scientists, policymakers, and the general public alike. Two major sets of factors have dominated studies attempting to predict democratization. One set characterizes endogenous or internal features of countries, and may be referred to as socioeconomic development. The other set, less often tested, characterizes exogenous variables that influence democratization via forces at work globally and within the region in which a country resides; this set may be referred to as diffusion processes. This study provides the first systematic comparison of these two sets of variables. When assessed alone, development indicators are robust predictors of democracy, but their predictive power fades with the inclusion of diffusion variables. In particular diffusion predictors of spatial proximity and networks are robust predictors of democratic growth in both the world and across all regions. The results demonstrate that regional patterns in democratization are evident, and hence world analyses are only the first approximation to understanding democratic growth. Finally, this study introduces an application of Multilevel Regression Models to studies on democratization. Such models fit observed data on world democratization better than the simple regression models used in most previous studies.